10. Nothing Certain But Death and Taxes: The Foster and Ellis Families

For more on the families mentioned in this post, please see The Omnibus Ancestry, available for download through Lulu.com.

It’s probably fair to say that in the world of genealogical sources, death and tax records are among the most underused. For those of us who have made the “trip to the courthouse” (increasingly becoming “a few hours online”), a day’s worth of research has often consisted of a check of marriage records to establish husband-to-wife connections, a run through a county’s estate records to establish generation-to-generation connections, and time permitting, a search of deeds to extablish land ownership and its progression from one family member to the next. Tax records, in contrast, are cumbersome if they are available at all, as they are generally unindexed. Depending on the state, death records might not even be in the county courthouse, but in a health department or city depository off site. Both are frequently skipped given limited time, in the interest of maximizing genealogical yield.

Nevertheless, even if seeking out tax and death records can be cumbersome, the rewards can be huge. To explain why, I present case studies within the Foster and Ellis families.

Tax Records: The Fosters of Albemarle County, Virginia

For many years our earliest known Foster ancestor was William Foster, who married Margaret Buster in Albemarle county, Virginia, in 1795 [1]. Although my father and I found several land transactions in the county under his name dating from the early 19th century, there appeared to be more than one person of the name and so we could not be sure which deeds pertained to our William [2]. It was clear, however, that he removed to Johnson county, Indiana, by 1821 [1]. Earlier generations, in contrast, were completely unclear. For a long time our working hypothesis was that his father was Edmond Foster, who figured repeatedly in Albemarle county deeds in the 1790s [3].

Thus the problem remained for about 20 years, unsolved and seemingly unsolvable. Finally in desperation, no other records seeming to be available, I accessed a microfilm of Albemarle land tax records through interlibrary loan. My expectation was that I might at least be able to sort out which land transactions applied to our William. That turned out to be true: In 1796, the year after his marriage, William was taxed on 288 acres. It was not until 1807 that a second William Foster appeared in the records [1,2]. Since we knew our ancestor had been in Albemarle throughout the intervening years, he had to have been the owner of the 288 acres.

But what I found next surprised me. It was a notation I never would have expected to find in a tax record. There in black and white, under the year 1807, were the 288 acres listed under the name “William Foster (James son)” [1].

In retrospect, it made sense. When another William Foster came on the scene, the tax collector needed to distinguish between them. It was our great good fortune that he had chosen to name my ancestor’s father. Edmond was out; James was in.

Land tax records having proven their value, I later sought out personal tax records. These also proved important, if not quite as spectacularly. Our William Foster disappeared from them after 1807 in Albemarle county [1], perfectly consistent with his migration to Nelson county, Kentucky, where he was found to be listed in the 1810 census before eventually removing to Indiana [1]. That sugggested the solution to another mystery, namely why he had continued to be taxed on the 288 acres through 1814 [1]. He had certainly departed, as attested by the personal tax records, but he had retained the property in his absence. Perhaps he left it in the care of a relative or tenant. We soon found a deed dated 1815, selling the bulk of the land while naming him as a resident of Nelson county [1].

Tax records in fact have proved a gift that continues giving with respect to the Foster family. While working out James Foster’s origins it became clear that he was one of several of the surname in St. Martin’s parish in Louisa county, Virginia. But there was also a group of Fosters in Trinity parish, Louisa county [1]. Were they related? The answer became apparent only by way of tax records. In 1767, our James was taxed in St. Martin’s parish on a slave named Constable. Then in 1775, Thomas Foster of the Trinity parish group, was taxed in St. Martin’s parish, also on a slave named Constable [1]. That was too big a coincidence to be coincidence; surely James had sold Constable to a family member. A tax collector’s diligence, even though it required combination with the great wrong of slavery, had tied together two halves of the Foster family.

Death Records: The Ellis Family of Johnson County, Indiana

In most U.S. jurisdictions, death records are a relatively recent innovation. They may not start until the late 1800s, and early records are often spotty. Then too, information in early death records may be limited to the date and place of death. More useful information, from a genealogical standpoint, often postdates the establishment of death records by a number of years.

Nevertheless in spite of frequent difficulty in accessing them, and the limited information they sometimes convey, death records are sometimes of great benefit. A good example comes from one of my two Ellis lines. For years my father was stuck in identifying the parents of ancestor Jesse Ellis (born 1828) of Johnson county, Indiana. Dad knew from census records that in 1870 and 1880, Jesse’s mother Leah was living with him, but that was all: We had no maiden name, and no identification of her husband [3].

Ellis, Jesse & Elizab.blog

Death records proved key. An index search revealed that Leah died in 1884, a scant six years after the beginning of death records in Johnson county. Her son Jesse’s death was also found in the index, dated 1906 [1,4]. A query to the health office in the county seat of Franklin produced critically important information in both cases. In her death record, Leah’s parents were shown as unknown — but her maiden name as Holman. And Jesse, who was already known from a short biography in a county history to have been born in Brown county, Ohio, was stated to have been the son of a Robert Ellis [5] .

These fragments of information were enough to solve the mystery of Jesse’s origins. In Brown county, Ohio, Dad located the 1819 marriage record of Robert Ellis, which in an early transcription gave his wife’s name as Sarah Holaman [6]. That was a close enough match that even though his wife’s name did not appear to be Leah, an estate record search was triggered, revealing that while administration of his estate was assigned to Sarah Ellis, she in fact signed documents as Leah Ellis [4]. The marriage record therefore clearly concerned Jesse Ellis’ parents. With this information in hand, we were able to trace both sides of the family for multiple generations, including descents from the Brasseur, Burckhart, Heilman, Rudolph, Str√∂hlin, and Veatch families [1].

It is true, though, that we likely would have solved this problem eventually by other means. A subsequent check of the manuscript marriage record showed the bride’s name to have been Leah Hollman, not Sarah Holaman [7]. That would have come to light at some point. With the knowledge that Jesse Ellis was born in Brown county, Ohio, to a mother named Leah, the relevance would have been clear. Nevertheless, making the connection through death records advanced research on the family by a number of years.


Like all sources, tax and death records have weaknesses that accompany their strengths. Notwithstanding the Foster example, tax records often do not clearly differentiate between taxpayers of the same name. This can lead to outright conflation. For example, in the Albemarle county land tax record of 1806, William Foster was shown as owning 320 acres. That was a conflation of our William, who as we have seen owned 288 acres, and another William who owned 32 acres [1]. Perhaps it was recognition of this error that in 1807 led the tax collector to separate the entries and to our great good fortune, label the larger tract as belonging to “William Foster (James son)”. A second weakness is also apparent from this example: The critical record may appear in only a single year’s listing. Thus a tax record search should be exhaustive if all available clues are to be had, and that can be very time-consuming, painfully detailed work.

In spite of these weaknesses, a strength of tax records is that they are contemporaneous. They are typically made yearly and thus contain important information recorded as it happened. This can be a great advantage in states with clearly defined “tithables”, as in post-colonial Virginia. A tithable was a male of a certain age, subject to tax, and so year-to-year changes in the number of tithables can offer valuable clues as to the ages of sons within a family.

Death records, in sharp contrast, are anything but contemporaneous except for information on the death itself. Statements of parentage or birth date and place rely not on the deceased, but on those left behind. If family members have no records, and rely only on all-too-human memory, the information can be wildly inaccurate. Thus whenever possible, death record information should be corroborated through other sources.

For both kinds of evidence, the only certainty is the inevitability of death and taxes. Perhaps only in the genealogy world is that good news.

For more on the families mentioned in this post, please see The Omnibus Ancestry, available for download through Lulu.com.


[1] Boles, D.B. (2016). The Omnibus Ancestry. E-book available for download through Lulu.

[2] Boles, H.W., & Boles, D.B. (1990). Foster Ancestors: Some Europeans, Immigrants, Colonists, and Pioneers. Decorah, Iowa: The Anundsen Publishing Co. Available for download through Lulu.

[3] Boles, H.W., & Boles, D.B. (1970). Some Earlier Americans: Boles and Bowers Relatives. Kalamazoo, Mich: private print, 1st edition.

[4] Boles, H.W., & Boles, D.B. (1994). Ellis Ancestors: Some Immigrants, Colonists, and Pioneers. Kalamazoo, Mich: private print. Available for download through Lulu.

[5] Letter from the Johnson County, Indiana, Health Office (1970).

[6] Notes of Harold W. Boles, original source unknown.

[7] FamilySearch, LDS Church.